v0.2.0

[0.2.0] - 2026-05-31

Added

• vona-core generation contracts for text token streaming plus decoupled audio transcription and synthesis traits.
• vona-ollama loopback text adapter for Ollama /api/generate streaming, exposed through the umbrella vona/ollama feature.
• vona-mlx Apple Silicon MLX audio engine facade, optional mlx-models text-loader hook, and umbrella MLX feature exports.
• vona-mlx-speech shared native Rust speech model loading utilities for safetensors discovery, metadata parsing, and MLX tensor materialization.
• vona-mlx-whisper native MLX Whisper loader and smoke example surface for local speech recognition experiments.
• vona-mlx-qwen3-tts native MLX Qwen3 TTS loader and smoke example surface for local speech synthesis.
• Vona model-provisioning manifests for Distil-Whisper Large V3 and the bf16 MLX Qwen3 TTS checkpoint so downstream apps can explicitly download speech assets before loading vona-mlx.
• mlx_ollama_voice_bench example and MLX/Ollama benchmark documentation covering the local voice pipeline validation flow.

Changed

• Split the MLX backend surface into granular contracts and crates so Ollama text generation, MLX audio orchestration, shared speech loading, Whisper, and Qwen3 TTS remain independently optional.
• Split vona-mlx feature gates so native-mlx enables only the MLX runtime and mlx-models-loader opts into the heavier mlx-models and tokenizer text-loader stack.
• Narrowed workspace reqwest defaults so local Ollama-only builds do not inherit TLS support, while remote HTTPS crates opt into Rustls explicitly.
• Removed unused direct dependency surface from vona-mlx after the bloat review without changing runtime behavior.

Fixed

• Corrected native Qwen3 TTS vocoder transposed-convolution stride inference so 12.5 Hz speech tokens decode to the expected 24 kHz waveform duration.
• Corrected the Ollama Phi mini default tag to phi4-mini, matching the Ollama library tag accepted by ollama pull.
• Updated benchmark notes to mark the earlier long-run Qwen3 TTS timings as historical because they were collected before the stride fix.

v0.1.1

[0.1.1] - 2026-05-16

Added

• vona-qwen crate with Qwen realtime ASR/TTS configuration helpers and event mapping utilities.

v0.1.0

0.1.0 - 2026-05-14

Added

• Initial Vona speech-to-speech runtime contracts, session driver, runtime policy, skill registry, passthrough backend, Seamless adapters, Moshi adapter surface, local transport helpers, sidecar, and deterministic test harness.
• Cloud provider protocol crates for OpenAI Realtime, Gemini Live, Azure Voice Live/Speech, ElevenLabs, and Deepgram.
• vona-model-provisioning for local model manifests, cache planning, and explicit HTTP artifact provisioning.
• Umbrella vona features for cloud adapters and local model provisioning.
• Release automation for version bumping, changelog maintenance, package checks, and publish-order crates.io releases.
• Integrity checks for model provisioning artifacts using expected size and SHA-256 metadata.

Changed

• Split the facade crate so vona-core owns provider-neutral runtime contracts while vona re-exports stable public surfaces and optional adapters.
• Expanded the release gate to cover provider crates and model provisioning.
• Remote Seamless sidecar responses now preserve all output audio frames while retaining the original single-frame fields for compatibility.

Fixed

• CI clippy compatibility for newer Rust toolchains in the Moshi WebSocket receive loop.
• Hosted realtime response completion is no longer treated as session closure.